Use of Genetics Algorithm in Estimating Bearing Capacity of Pile under Dynamic Load

Pile foundations provide support for structures, transferring their load to layers of soil or rock that have sufficient bearing capacity and suitable settlement characteristics. There are a very wide range of foundation types available, suitable for different applications, depending on considerations. There are various methods to estimate the axial bearing capacity of the pile such as experimental, analytical and in-situ experiments. Accurate estimation of bearing capacity is essential in geotechnical engineering due to the high cost of implementing deep foundations, that use of artificial intelligence and genetic algorithms can be helpful in this regard. In this study compressive bearing capacity for two prefabricated concrete and steel pipe evaluated under harmonic load using modeling in finite element software (Plaxis 2D), and the results were compared with the Dynamic Load Testing (PDA) data in a case study. In this research, the results of 100 numerical analyzes have been optimized using a genetic algorithm that resulted in the introduction of an estimated relationship with the accuracy of the pile bearing axial bearing capacity for concrete and metal piles. In this study, comparing different methods of assessing the bearing capacity of piles indicates the fact that most empirical methods provide a much larger load capacity compared to actual values.